Creative Thinking in an Emotional Context: Specific Relevance of Executive Control of Emotion-Laden Representations in the Inventiveness in Generating Alternative Appraisals of Negative Events

Abstract

Although divergent thinking ability in different domains may largely rely on the same basic executive functions, domain-specific functions may also be important, in particular when it comes to more real-life creativity demands. This study investigated if functional executive control of emotion-laden representations may be specifically relevant in cognitive reappraisal, which implies being creative in an affective context. In a sample of 88 healthy individuals, the relation between the participants’ inventiveness in generating positive reappraisals of adverse events (Reappraisal Inventiveness Test) and in generating novel ideas without emotional component (conventional divergent thinking test) to their executive functioning in tasks without (Mittenecker Pointing Test) and with emotional contribution (humor processing task) was studied. In line with hybrid models of creative thinking, poorer basic inhibition skills were found to be associated with poorer fluency performance in both divergent thinking tasks. Relations applied more specifically to reappraisal inventiveness when it came to executive processes with a more prominent emotional component. Creative performance in both tasks may have been hampered by time limits. The results support the notion that, in addition to basic executive functioning, more specific cognitive control functions are implicated in more real-life creative performance, according to related domain-specific demands.

Creativity is a skill relevant in many contexts of everyday life, such as school, work, arts, sports, and coping with adversities (Hong & Milgram, 2010; Kaufman & Baer, 2004; Rawlings & Locarnini, 2007; Weber, Loureiro De Assunção, Martin, Westmeyer, & Geisler, 2014). Accordingly, the importance of investigating complex and more real-life creative behavior is increasingly recognized, which has also raised the question to which extent domain-general and domain-specific cognitive control processes are implicated in creative thinking (Agnoli, Corazza, & Runco, 2016; Baer, 1998; Boccia, Piccardi, Palermo, Nori, & Palmiero, 2015; Ellamil, Dobson, Beeman, & Christoff, 2012; Fink et al., 2018b; Fink, Perchtold, & Rominger, 2018a; Hong & Milgram, 2010; Palmiero, Nakatani, Raver, Belardinelli, & Van Leeuwen, 2010; Rominger et al., in press; Rominger, Reitinger, Seyfried, Schneckenleitner, & Fink, 2017b; for examples of domain-specificity of executive control processes see Cocchini, Logie, Della Sala, MacPherson, & Baddeley, 2002; Hamilton & Martin, 2005; Vuong & Martin, 2014).

A suitable yet still understudied creative skill for investigating the relevance of basic and domain-specific executive functioning in creative or divergent thinking is the inventiveness in generating alternative appraisals of negative events (Fink et al., 2017; Weber et al., 2014). It requires the generation of manifold alternative, original, and useful solutions to an open and ill-defined problem, matching the main characteristics of creativity (Runco & Acar, 2012; Runco & Jaeger, 2012). Thus, the generation of alternative appraisals of negative events implies being creative in an affective context (Fink et al., 2018a, 2017; Perchtold et al., 2018b; Weber et al., 2014).

Creative ideation is commonly thought to rely on a range of executive control functions (for an overview, see Fink & Benedek, 2014; Fink et al., 2018a), and their importance is recognized in prominent creativity theories (Allen & Thomas, 2011; Chrysikou, Weber, & Thompson-Schill, 2014; Finke, Ward, & Smith, 1996; Mok, 2014; Sowden, Pringle, & Gabora, 2015). Primarily top-down control processes, preservation of internal attention, and inhibition of habitual responses are regarded vital ingredients of creativity (Benedek, Franz, Heene, & Neubauer, 2012; Benedek, Jauk, Sommer, Arendasy, & Neubauer, 2014; Cheng, Hu, Jia, & Runco, 2016; Edl, Benedek, Papousek, Weiss, & Fink, 2014; Fink & Benedek, 2014; Pinho, Ullen, Castelo-Branco, Fransson, & De Manzano, 2016; Rominger, Fink, Weiss, Bosch, & Papousek, 2017a; Rominger et al., in press; Zabelina & Robinson, 2010; Zabelina, Robinson, Council, & Bresin, 2012). Although a contribution of executive functioning has also been assumed in cognitive reappraisal (Ochsner & Gross, 2005; Weber et al., 2014), direct behavioral evidence is relatively scarce to date. It was proposed that the generation of reappraisals of stressful situations requires the ability to inhibit the prepotent negative aspect of the situation, along with switching (shifting focus) between negative and neutral or positive mental sets, and updating the current perspective with a newer one. Consequently, inhibition, set-shifting, and possibly memory updating should represent central control functions (Malooly, Genet, & Siemer, 2013). So far, the inventiveness in generating reappraisals has not been investigated in this context. Indirect information was derived from investigations studying the effectiveness of reappraisal efforts in terms of subjectively reported reduced negative affect in laboratory tasks. However, results were heterogeneous. Although some studies found positive correlations with memory updating performance (Hendricks & Buchanan, 2016; Schmeichel, Volokhov, & Demaree, 2008), others reported a counterintuitive negative association with set-shifting performance (McRae, Jacobs, Ray, John, & Gross, 2012), and no correlations with inhibition in the Stroop and a stop-signal task (Hendricks & Buchanan, 2016; McRae et al., 2012).¹

It has also been proposed that reappraisal ability may be related to executive control over emotional material, over and above the functionality of nonemotional control processes (Malooly et al., 2013). This notion was supported by direct comparisons of brain activity changes during the generation of ideas in a conventional divergent thinking task and a cognitive reappraisal task, which suggested that some additional demands must be implicated in cognitive reappraisal compared to divergent thinking without emotional component (Fink et al., 2017; Papousek et al., 2017; Perchtold et al., 2018b). These demands are presumably attributed to the linkup of participating executive processes with affective contexts. Taken together, this background suggested the investigation of executive functioning including basic, nonemotional functions as well as the functioning of cognitive control processes embedded in an emotional context.

One possibility to put relevant executive functions in an emotional context is to study humor processing. Jokes typically have a first, contextually salient interpretation that is discarded at the punch line in favor of a more marked reading (Brône, Feyaerts, & Veale, 2006; Giora, 1991). In typical jokes, this involves the inhibition of the negative emotional aspect of the situation (i.e., what is happening to the victim of the joke) and a shift of focus toward positive aspects or interpretations (Martin & Lefcourt, 1983; Moran, 1996) resembling putative demands in cognitive reappraisal.² Fortunately, there is a validated task (Lackner et al., 2013; Papousek et al., 2013a, 2013b; Papousek, Schulter, Lackner, Samson, & Freudenthaler, 2014; Samson, Zysset, & Huber, 2008) that allows one to observe two interdependent but dissociable processes implicating inhibition and cognitive shifting that increasingly involve the linkup of cognitive and emotional components (Gardner, Ling, Flamm, & Silverman, 1975; Shammi & Stuss, 2003; Vrticka, Black, & Reiss, 2013). The functionality of the more cognitive component of perceiving the punch line (i.e., of detecting and resolving the incompatible elements in the joke) is mirrored in the time until the punch line is detected (response latency). An even greater emotional share is implicated in the immediate experience of pleasure related to the satisfaction derived from the cognitive event of sudden insight when recognizing the punch line (Vrticka et al., 2013). It can be objectively quantified by analyzing the cardiac response to the detection of the punch line (contrasted transient response; Lackner et al., 2013). Links between divergent thinking and humor processing have long been recognized in the creativity literature (e.g., Cundall, 2007; Kellner & Benedek, 2017; O’Quin & Derks, 2011).

The appropriate measurement of executive functioning in nonemotional contexts constitutes another challenge altogether. It has been posited that some executive functions are difficult to quantify experimentally, because they are rooted in the temporal domain (Paulus, Geyer, & Braff, 1999; Zabelina & Robinson, 2010). This particularly applies to measures referring to the function of dynamically activating and modifying cognitive processes in response to changing conditions and demands, such as the inhibition of dominant responses and developing routines (perseveration) and memory updating. A well-designed motor random generation test allows one to separately measure individual differences in these two functions (Schulter, Mittenecker, & Papousek, 2010; see also Miyake et al., 2000). Several studies confirmed that the interindividual variability in the performance on such a test meaningfully relates to the efficiency of executive processes in healthy individuals (Schulter et al., 2010; Weiss et al., 2017, 2014). Previous studies using random generation tasks predominantly indicated the relevance of inhibition for fluency in conventional divergent thinking (Benedek et al., 2012; Cheng et al., 2016; Zabelina et al., 2012).

Taken together, the main aim of the study was to investigate if divergent thinking in an emotional context specifically requires executive control of emotion-laden representations that goes beyond basic executive functions (that are required in all divergent thinking). To this end, we evaluated the relevance of basic and domain-specific executive functioning in the inventiveness in generating alternative appraisals of negative events, which implicates being creative in an affective context. For a more integrated view, performance on the same executive tasks was also correlated with the inventiveness in generating novel ideas without emotional component (i.e., conventional verbal divergent thinking). It was expected that poorer inventiveness in the emotional context, as well as poorer inventiveness in the conventional divergent thinking task without emotional component, would be associated with poorer basic inhibition and possibly poorer memory updating skills, reflecting shared demands of the two types of divergent thinking. It was further expected that poorer inhibition and shifting in an emotional context (i.e., slower detection of punch lines and a less pronounced immediate pleasure response after the experience of sudden insight during humor processing) would be associated with poorer inventiveness in an emotional context (generation of alternative appraisals) over and above conventional divergent thinking ability.

Method

Participants

The sample comprised n = 88 university students (38 men, 50 women) who were enrolled in various fields. Their age was between 18 and 33 years (M = 23.1, SD = 3.5). Individuals with major psychiatric disorders/history of major psychiatric disorders according to the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID) and individuals who reported having a neurological disease or using psychoactive medication were not included in the study. A student sample was chosen because part of the standard version of the reappraisal test had been tailored for a student population (the items include negative experiences that students can easily imagine happening to them). The study was approved by the authorised ethics committee. Written consent was obtained from all participants.

Assessments

Divergent thinking in an emotional context: inventiveness in generating alternative appraisals of negative events

For the assessment of this specific divergent thinking ability, the Reappraisal Inventiveness Test (RIT; Weber et al., 2014) was used. This test was applied in several previous studies, which confirmed its validity and substantiated the notion of cognitive reappraisal being an instance of creative or divergent thinking (Fink et al., 2017; Papousek et al., 2017; Perchtold et al., 2018a, 2018b; Weber et al., 2014). The original RIT consists of four anger-eliciting vignettes. Each vignette is supplemented by a matching photograph to make the situation more vivid. For a broader assessment, four additionally developed and validated anxiety-eliciting vignettes (De Assuncao, Golke, Geisler, & Weber, 2015) were also used. Participants were instructed to imagine the situation happening to them and to generate and write down as many and as different ways as possible to think about or appraise the situation in a way that diminishes their anger or anxiety (3 min each). On average, participants generated 5.35 (SD = 1.31) valid, distinguishable reappraisals to each of the eight vignettes. The scores obtained for each vignette were aggregated to obtain the total test score (M = 42.81, SD = 10.48). Aggregation across anger and anxiety vignettes is justified by confirmatory factor analysis, which showed that reappraisal inventiveness should be regarded as a trans-emotional ability that is not specific to certain emotions (De Assuncao et al., 2015). Accordingly, the internal consistency reliability across the eight anger and anxiety items (Cronbach’s alpha) was α = .91 (similar reliabilities were reported in Papousek et al., 2017; Weber et al., 2014). Total numbers of reappraisals in the three categories that are available in the category schemes for both the anger and the anxiety vignettes were also calculated (see Weber et al., 2014): positive reinterpretation (generating positive aspects; α = .74, M = 16.5, SD = 7.1), problem-oriented (casting the situation in terms of how the impact of the situation could be reduced or compensated for; α = .70, M = 7.2, SD = 4.1), and de-emphasising (α = .72, M = 17.2, SD = 7.0). The fluency index was scored by an experienced researcher. Papousek et al. (2017) and Weber et al. (2014) reported very satisfying interrater reliabilities of around ICC = .90. In addition to this fluency score the RIT also allows for the calculation of a flexibility score (number of categorically different reappraisals). However, in the RIT, fluency and the flexibility scores are highly redundant (r = .90 in our study; r = .97 in the study of Weber et al., 2014). Therefore, the analyses focus on the fluency score only. The RIT does not provide for the determination of the originality of reappraisal ideas.

Conventional divergent thinking: inventiveness in generating novel ideas without emotional component

For the assessment of the capacity to generate novel ideas without emotional component, the verbal imagination subscales of the German Berlin Intelligence-Structure Test (Jäger, Süß, & Beauducel, 1997) were used, which capture divergent thinking applied to verbal content. The four verbal subtests require participants to produce and write down as many different ideas as possible (e.g., alternative uses, insights) in a limited amount of time (2 to 2.5 min). Because it was the study’s goal to relate the findings for the inventiveness in generating ideas with to those without emotional component, in analogy to the RIT only the fluency score of the BIS scales was used in the analyses (total number of generated ideas; α = .62, M = 31.4, SD = 7.4).

Executive functioning in an emotional context: humor processing task

A well-validated nonverbal humor-processing task was used for the assessment of executive functioning in an emotional context (Lackner et al., 2013; Papousek et al., 2013a, 2014, 2013b; Samson, Lackner, Weiss, & Papousek, 2012; Samson et al., 2008). The computer-based task comprised 32 humorous cartoons and 16 nonhumourous control pictures (all nonverbal line drawings). All humorous cartoons contained an incongruity which could be resolved meaningfully (i.e., they involved a typical punch line) and followed a typical joke pattern (i.e., involved a target of the joke). Control pictures contained incongruities that could not be resolved and typically are not perceived funny. Please see Samson et al. (2012) for examples of the humorous material. In our study, only the more complex cartoons were used (semantic and theory of mind cartoons), which implicate some emotionally toned incongruity. The more simple puns, in which the incongruity is typically nonemotional, and for which the cardiac contrasted transient response is difficult to determine because of very short response times, were not used. Participants were instructed to indicate via mouse click whether they had or had not understood the punch line and to click the respective button as soon as they had understood the punch line or were sure that they did not understand it. The pictures were presented in random order until the understood or not understood button was clicked, respectively (max. 6 s), followed by max. 4 s in which the picture’s funniness was rated on a scale ranging from 1 to 6, 10 s in which the punch line should be briefly described and another 10 s in which a fixation cross was presented, before the next picture appeared on the screen. On average, 24.9 (SD = 4.5) or 77.81% of the punch lines were detected. Participants erroneously indicated having perceived a punch line in M = 3.2 (SD = 2.3) or 19.96% of the nonhumorous control pictures (false alarms; d’ = 1.61). The funniness of cartoons was rated M = 3.5 (SD = 0.9; control pictures: M = 1.3, SD = 0.4). These numbers confirm that participants understood the task and that the two relevant measures of the task (i.e., response latencies and the cardiac Contrasted Transient Response) are based on an adequate data basis.

Response latencies (starting from the onset of the pictures) were averaged across all detected punch lines in cartoons (M = 4.0, SD = 0.49, min = 2.6, max = 4.9). Residualized scores were calculated, to control for individual differences in general response tendencies and response speed, for analyzing variability specific to the detection of humorous punch lines. This was done by conducting linear regressions using the scores of the (not understood) control pictures to predict those of the (understood) cartoons (see also Samson et al., 2012). Shorter response latencies indicate higher efficiency of the moderately emotion-laden cognitive process of detecting the humor, which implicates the functionality of inhibition and set-shifting in an affective context (Martin & Lefcourt, 1983; Moran, 1996).

Cardiac contrasted transient response (CTR; Lackner et al., 2013) data were obtained with electrocardiogram recording during the humor-processing task using a standard limb lead II electrode configuration. Heart rate changes relative to the 0.5 s frame preceding the picture onset were separately averaged across trials with cartoons and trials with nonhumorous control pictures, and the CTR was calculated as the difference between the heart rate change in the 0.5 s frame immediately preceding the participant’s response indicating having detected the punch line in cartoons minus the data from the same 0.5 s frame in control trials in which participants did not find a punch line (M = 1.3 bpm, SD = 2.4). Higher positive values indicate more immediate pleasure related to the satisfaction derived from the sudden insight when detecting the humor. Thus, while referring to the same cognitive process of detecting and resolving the incompatible elements in the joke, in the CTR even more emotional load is implicated compared to the response latency. Lackner et al. (2013) presented details for calculating the CTR and its validity.

Executive functioning without emotional contribution: mittenecker pointing test

The Mittenecker Pointing Test (MPT; Schulter et al., 2010) was selected as a presumably particularly appropriate executive test in this context. It is a computer-based motor random generation test in which the participants are instructed to press the keys of a keyboard with nine unlabelled keys irregularly distributed over the board with their right index finger in the most random or chaotic order possible (180 responses in total). The responses were paced by an acoustic signal (1.2/s) to control the rate of production. The MPT allows to separately measure individual differences in two important executive functions with the same simple test by using sophisticated and validated quantitative measures of deviation from randomness, based on information theory analysis (Schulter et al., 2010): The symbol redundancy (SR) is related to memory monitoring and updating. A SR score of zero denotes maximal equality of the relative frequencies of chosen keys and thus minimal predictability, whereas a score of 1.0 denotes maximal redundancy and, thus, a lack of randomness (i.e., higher values indicate poorer performance; M = .011, SD = .009). The context redundancy (CR) is based on the sequential probability of each chosen key. The major part of the interindividual variance in CR is due to the tendency to repeat certain response sequences en bloc. Thus, CR reflects the inhibition of developing routines. A CR score of zero denotes the complete absence of any regular pattern; a score of 1.0 denotes the presence of a fixed, repetitive response pattern (i.e., higher values indicate poorer performance; M = .18, SD = .04). For detailed information on the test and how to compute SR and CR please see Schulter et al. (2010).

Procedure

After filling in the consent form, participants completed the RIT anxiety items, the BIS subtests, the MPT, and the RIT anger items. One to 3 weeks later, they completed the SCID and the humor-processing task. All tasks were administered by an experienced researcher in a quiet examination room located at the department.

Statistical Analysis

The main aim of the study was to investigate if divergent thinking in an emotional context specifically requires executive control of emotion-laden representations, beyond basic executive functions that are required in all divergent thinking. Hence, in a first step, Pearson correlations (r) were used to test if poorer performance on the RIT (divergent thinking in an emotional context) as well as on the BIS (conventional divergent thinking) were correlated with poorer basic executive functioning (executive functioning without emotional contribution: inhibition and memory updating in the MPT). The evaluation whether the executive functions in the MPT reflect basic requirements that are relevant in all divergent thinking irrespective of the specific context (emotional or not; first hypothesis) corresponds to the statistical question whether it is shared variance of RIT and BIS performance that accounts for their correlations with executive functioning in the MPT. Therefore, significant Pearson correlations were followed-up by standard multiple regression analysis with performance on the RIT and performance on the BIS as independent variables and the respective component of the MPT as the dependent variable. The Pearson correlations include variance that RIT and BIS performances share, as well as unique portions of variance of RIT and BIS performance that are associated with the MPT component. The semipartial correlations (sr), obtained in the multiple regression analysis, contain the unique portions of variance only. Hence, if the semipartial correlations are lower than the Pearson correlations (and nonsignificant), it can be concluded that predominantly shared variance of RIT and BIS performance accounts for their correlations with the MPT component. A significant multiple regression coefficient (R) in spite of nonsignificant semipartial correlations of the individual independent variables in the model corroborates this conclusion.

To evaluate whether executive control of emotion-laden representations (obtained in the humor processing task) is specifically associated with divergent thinking in an emotional context (RIT), over and above conventional divergent thinking ability (BIS; second hypothesis), an analogous statistical strategy was applied. A significant Pearson correlation of RIT performance with one of the emotion-laden executive components was followed-up by a standard multiple regression analysis with performance on the RIT and performance on the BIS as independent variables and the respective emotion-laden executive component as the dependent variable. If the association between RIT performance and an executive function indicated in the Pearson correlation is preserved in the respective semipartial correlation, it can be concluded that specific variance of RIT performance, that is, variance not shared with conventional divergent thinking ability, decisively accounts for its correlation with the respective executive function. In this case, the respective executive function is specifically relevant to divergent thinking in an emotional context (i.e., to the inventiveness in generating alternative appraisals of negative events), beyond conventional divergent thinking ability.

Together, these analyses evaluate the importance of executive functions with increasing emotional contribution, starting from basic executive functioning without emotional contribution (MPT), followed by the functionality of executive processes with some emotional contribution (response times in the detection of humor), and finally the indicator implicating the greatest emotional load (immediate pleasure response after detecting the humor, CTR).

Results

Pearson correlation analyses indicated that poorer inhibition of prepotent responses in the MPT was associated with poorer inventiveness in generating alternative appraisals of negative emotional events (CR; r = –.23, p = .032) as well as with poorer performance on the non-emotional divergent thinking test (CR: r = –.24, p = .025). Memory monitoring and updating skills were not significantly related to reappraisal performance (SR; r = –.18, p = .097), nor to the performance on the non-emotional divergent thinking test (SR: r = –.03, p = .815). The low semipartial correlations in the multiple regression analysis revealed that the inhibition of prepotent responses was mainly correlated with shared variance of the inventiveness in generating positive reappraisals and divergent thinking without emotional component (RIT: sr = –.11, p = .325; BIS: sr = –.13, p = .243; R = .26, p = .050). The intercorrelation between RIT and BIS fluency was r = .62 (p < .001).

Poorer inventiveness in generating positive reappraisals was related to slower detection of punch lines in jokes involving also negative emotional aspects (r = –.24, p = .026). Performance on the BIS was not significantly correlated to this emotionally toned cognitive process of humor perception (r = –.17, p = .106). In the multiple regression analysis, the semipartial correlation of reappraisal inventiveness with the response latencies was pushed under the significance threshold (sr = –.17, p = .119), indicating the contribution of some shared variance of reappraisal inventiveness and conventional divergent thinking in this relationship (BIS: sr = –.03, p = .750).

Finally, the process with the greatest emotional component, the occurrence of immediate pleasure after perception of the punch lines in the jokes, was related to reappraisal inventiveness. Individuals showing a less pronounced CTR generated a smaller number of positive reappraisals in the RIT (r = .26, p = .016). The CTR was not associated with performance on the BIS (r = .09, p = .403). In the multiple regression analysis, the semipartial correlation of reappraisal inventiveness remained unchanged compared to its zero-order correlation (sr = .26, p = .016), indicating that shared variance between the two abilities did not play a role in this case (BIS: sr = .09, p = .405).

Additional analyses of the types of generated reappraisals showed that the significant correlations were mainly driven by reappraisals in the category positive reinterpretation. Table 1 summarizes these correlations. Intercorrelations showed that largely different processes were represented by the inhibition component of the MPT, the response latency, and the cardiac response in the humor processing task (CR × humor response latency: r = .08, p = .451; CR × humor cardiac response: r = –.18, p = .093; humor response latency × cardiac response: r = –.04, p = .703).

TABLE 1 Summary of correlations between the inventiveness in generating alternative appraisals of negative events (RIT) and the inventiveness in generating novel ideas without emotional component (BIS) and executive functioning with increasing emotional contribution

	MPT Inhibition (CR)		Humor Task Inhibition/Shifting (Response Latency)		Humor Task Inhibition/Shifting (CTR)
	r (p)	sr (p)	r (p)	sr (p)	r (p)	sr (p)
RIT total	−.23(.032)	−.11(.325)	−.24(.026)	−.17(.119)	.26 (.016)	.26 (.016)
RIT positive reinterpretation	−.27(.012)	−.17(.098)	−.25(.019)	−.19(.074)	.24 (.027)	.22 (.040)
RIT problem-oriented	.11 (.297)	.21 (.046)	−.05(.640)	.01 (.922)	−.01(.912)	−.05(.669)
RIT de-emphasising	−.15(.173)	−.10(.360)	−.03(.757)	.01 (.964)	.19 (.081)	.17 (.112)
BIS	−.24(.025)	−.13(.243)	−.17(.106)	−.03(.750)	.09 (.403)	.09 (.405)

Note: r … Pearson correlation, sr … semi-partial correlation controlling for performance on the BIS (first four rows)/the RIT (last row). Correlations with p < .05 are highlighted in bold font.

Discussion

The aim of this study was to investigate the relevance of basic and domain-specific executive functioning to an instance of real-life creative ability, that is, to the inventiveness in generating alternative appraisals of negative events, which implicates being creative in an affective context. The study showed that the inventiveness in generating cognitive reappraisals was related to the basic inhibition of prepotent responses and perseveration as well as to the functionality of control processes in the emotional context of humor processing. The findings indicated that poorer basic inhibition skills, as well as poorer inhibition and cognitive shifting in an emotional context, are linked to poorer reappraisal inventiveness. Taking account of conventional divergent thinking ability, relations applied more specifically to reappraisal inventiveness for functions with a more prominent emotional component.

The relevance of basic inhibition skills to the inventiveness in generating novel ideas with and without emotional component is in line with previous empirical findings and theoretical accounts in the field of creativity pointing to the importance of executive top-down control processes in creative ideation (Edl et al., 2014; Finke et al., 1996; Groborz & Necka, 2003; Lopata, Nowicki, & Joanisse, 2017; Pinho et al., 2016; Rominger et al., 2017a, in press; Zabelina & Robinson, 2010). The correlations directly corroborate previous findings that were obtained with the same or a similar random generation test and conventional divergent thinking ability (Benedek et al., 2012; Cheng et al., 2016). Our results add to these previous findings by providing behavioral evidence for the notion that functional basic inhibitory processes are important for creative ideation independently of the domain or context in which they are embedded. In line with that, inhibition was primarily related to shared variance of the inventiveness in generating alternative appraisals of negative events and conventional creative ideas.

Previous studies in the field of cognitive reappraisal did not find significant correlations with response inhibition in the Stroop and a stop signal task (Hendricks & Buchanan, 2016; McRae et al., 2012). The MPT, allowing to assess temporally rooted processes such as the inhibition of developing routines, may more closely map relevant processes. A further important difference with those previous studies is that they used subjective ratings of negative affect elicited by negative stimuli to assess the self-reported effectiveness of reappraisal efforts, which is only a very indirect indicator of the inventiveness in generating reappraisals and, hence, creative or divergent thinking. By contributing to the capability to switch to alternative appraisals, functional inhibition may primarily help to increase the pool of generated reinterpretations, that is, the inventiveness in generating reappraisals, which had not been directly related to inhibitory functions, so far.

In contrast to the inhibition component (CR), the MPT’s memory updating component (SR) was related to neither of the two divergent thinking tasks. The inhibition component refers to the ability to inhibit or override the tendency to produce a more dominant or automatic response when necessary for achieving the current behavioral goal (Miyake et al., 2000), and thus seems conceptually closer to the putative demands in creative thinking, compared to memory updating. Zabelina et al. (2012) did also not find a correlation between SR and conventional divergent thinking performance, and studies reported performance on conventional creativity tasks to be correlated with inhibition in random sequence generation (Benedek et al., 2012; Cheng et al., 2016). In line with our findings, Camarda et al. (in press) indicated the specific importance of inhibitory control (but not working memory) for creative ideation performance by means of an experimental dual task design. Studies in the domain of cognitive reappraisal reported positive correlations between reappraisal performance and memory updating skills, but these studies did not directly assess the inventiveness in generating cognitive reappraisals but effects of reappraisal efforts on mood (Hendricks & Buchanan, 2016; Schmeichel et al., 2008). The SR score of the MPT is a specific indicator of monitoring and updating contents in memory that essentially taps the individual’s capability to keep track of the chosen responses. The working memory load may be considered relatively low, which might mask subtle processing deficits related to memory monitoring/updating (Weiss et al., 2017). In line with this, several studies in clinical, as well as in healthy, samples indicated that CR is the more sensitive measure compared to SR (Schulter et al., 2010; Weiss et al., 2017).

Performances in the RIT and the conventional divergent thinking test shared about 35% of their variance, as was also shown by Weber et al. (2014) and Fink et al. (2017). Nevertheless, creativity does not seem to be as unitary as has been proposed by some authors (e.g., Julmi & Scherm, 2015; Plucker, 1998; Simonton, 2009; but see Boccia et al., 2015; Fink et al., 2018b; Hong & Milgram, 2010; Palmiero et al., 2010; Pidgeon et al., 2016; Pinho et al., 2016; Silvia, Kaufman, & Pretz, 2009). The functionality of the studied executive processes was associated with some unique variance of reappraisal inventiveness. This seemed to be the more true the more prominent the emotional component in the executive function was, corroborating evidence that the emotion-related features in cognitive reappraisal place additional demands on the brain compared to conventional divergent thinking (Fink et al., 2017; Papousek et al., 2017). Together, these findings shed some light on reports of weak or even zero correlations between creative performances in different domains (Baer, 1998; Baldo, Shimamura, Delis, Kramer, & Kaplan, 2001; Fink et al., 2018b; Kaufman & Baer, 2004; Rawlings & Locarnini, 2007; Regard, Strauss, & Knapp, 1982; Rossmann & Fink, 2010; Ruff, Light, & Evans, 1987; Runco, 1987; for a summary see Weisberg, 2006; but see Ulger, 2015). Correlations between different types of divergent thinking tasks may vary depending on the share of basic versus domain-specific executive functioning underlying adequate performance.

In this study, the emotional context for the operation of relevant cognitive control functions was established using a validated humor processing task. The detection and appreciation of the punch line in a typical joke requires functional executive control (inhibition and cognitive shifting) of emotion-laden representations (Martin & Lefcourt, 1983; Moran, 1996; Samson et al., 2012). Compared to the generation of problem-oriented appraisals or de-emphasising, the proposed contributing control processes apply most to positive re-interpretation. Consequently, the finding that the correlations of this study were mainly driven by the number of generated reappraisals categorized as positive reinterpretation enhances the plausibility of the pattern of results. In line with these findings, poorer shifting away from negative and toward positive emotional aspects in an emotional task-switching task (Malooly et al., 2013) was associated with less successful reappraisal efforts in terms of subjectively reported reduced negative affect. Please note that the correlation between reappraisal inventiveness and the response latencies in the humor task cannot be explained by individual differences in general processing speed, because residualized scores adjusted for such general tendencies were used (see methods section).

This study broke new ground in several ways. The study of the inventiveness in the generation of alternative appraisals of negative events is novel to the field of creativity. Studying creative ideation in the context of coping with adversities may constitute a fruitful avenue for further enhancing the practical relevance of creativity research (see also Reynolds, 2003, 2004; Reynolds, Lim, & Prior, 2008; Zausner, 1998). By exclusively drawing on objective behavioral measures instead of self-reported data, our data are independent from the participants’ ability or willingness to accurately report on their own experience. Moreover, the inventiveness in generating positive reappraisals and conventional creative ideas, the structure of generated random patterns, response latencies and transient heart rate responses are all different kinds of objective behavioral measures. Therefore, common methods variance and response biases are entirely excluded, which also puts the relatively low effect sizes into perspective. Beyond that, it is the first study directly relating reappraisal ability to control processes in the context of humor processing, and the first evaluating the relevance of executive functioning for the generation of different types of cognitive reappraisals.

A limitation of this study is that the time to respond was restricted in both divergent thinking tasks, which may have influenced the creative ideation process to some extent. However, although the originality of ideas may have been affected (Cheng et al., 2016), time limits are essential to reliably measure individual differences in fluency performance, which is the most important measure in the reappraisal task. A further limitation is the use of the fluency score as the only indicator for divergent thinking performance, due to the high redundancy of fluency and flexibility scores in the reappraisal task (see also Fink et al., 2017; Weber et al., 2014). Although an individual’s creative potential may be better estimated by use of multiple indices (Runco & Acar, 2012), the fluency of ideas qualifies as a valid indicator of creativity (e.g., Jauk, Benedek, & Neubauer, 2014).

In conclusion, these support the notion that unimpaired executive functioning plays a role in the inventiveness in generating novel ideas in general, as well as in the inventiveness in generating alternative appraisals of negative events in particular (cf. Fink & Benedek, 2014; Fink et al., 2018a; Malooly et al., 2013; Ochsner & Gross, 2005; Weber et al., 2014). Compared to conventional divergent thinking ability, additional executive demands seem to be implicated in reappraisal inventiveness, attributed to the linkup of cognitive processes with affective contexts (cf. Fink et al., 2017; Malooly et al., 2013; Papousek et al., 2017; Perchtold et al., 2018b). The findings are in line with hybrid theories of creativity assuming that some cognitive control functions impact creative performance across domains, whereas others may exert domain-specific effects (Boccia et al., 2015; Hong & Milgram, 2010; Julmi & Scherm, 2015; Kaufman & Baer, 2004,p. 9; Weisberg, 2006). Together, the results support the notion that in addition to basic executive functioning, more specific cognitive control functions are implicated in more real-life creative performance, according to related domain-specific demands.

Additional information

Funding

This work was supported by the Austrian Science Fund under Grant P 30362.

Notes

¹T he behaviorally assessed inventiveness in generating alternative appraisals of negative emotional events, on which this study focuses, is unrelated to the self-reported habitual use of reappraisal (Weber et al., 2014). Therefore, and for space limits, research having used self-reported habitual tendencies to engage in reappraisal efforts is not reviewed here.

² Depending on the type of joke additional, more complex cognitive processes such as mentalising or anticipation of events that are only insinuated may be required for understanding the punch line (Weiss et al., 2013).